Comparison of Annual Maximum and Peaks Over Threshold Methods with Automated Threshold Selection in Flood Frequency Analysis: A Case Study for Australia

Abstract

Flood frequency analysis (FFA) enables fitting of distribution functions to observed flow data for estimation of flood quantiles. Two main approaches, Annual Maximum (AM) and peaks-over-threshold (POT) are adopted for FFA. POT approach is under-employed due to its complexity and uncertainty associated with the threshold selection and independence criteria for selecting peak flows. This study evaluates the POT and AM approaches using data from 188 gauged stations in south-east Australia. POT approach adopted in this study applies a different average numbers of events per year fitted with Generalised Pareto (GP) distribution with an automated threshold detection method. The POT model extends its parametric approach to Maximum Likelihood Estimator (MLE) and Point Moment Weighted Unbiased (PMWU) method. Generalised Extreme Value (GEV) distribution using L-moment estimator is used for AM approach. It has been found that there is a large difference in design flood estimates between the AM and POT approaches for smaller average recurrence intervals (ARI), with a median difference of 25% for 1.01 year ARI and 5% for 50 and 100 years ARIs.

Estimation of Large River Design Floods Using the Peaks-Over-Threshold (POT) Method

“…Namely, during the last 20-30 years, floods have become more frequent, which is unlikely to affect the output of statistical analyses of long series. In their study, Pan and Rahman [26] illustrate the advantages of using the POT method rather than the annual maximum series for providing a better understanding of the nature of changes in flood events. Even if the annual maximum approach is used in most cases, it cannot represent the complexity of the flood regime [27].…”

Comparison of annual maximum and peaks-over-threshold methods with automated threshold selection in flood frequency analysis: a case study for Australia

“…Based on the purposes of the analysis, POT can retain infinitive data points above the selected threshold value provided that the selected data points meet the independence criteria. Many studies have suggested the advantages of using POT approach in FFA (Bezak, Brilly & Šraj 2014;Bhunya et al 2012;Bhuyan, Jena & Bhunya 2016;Bobée et al 1993;Gottschalk & Krasovskaia 2002;Herath, Prasad Basnayake & Coremans 2015;Karim, Hasan & Marvanek 2017;Madsen, Rasmussen & Rosbjerg 1997;Mostofi Zadeh et al 2019;Pan & Rahman 2018). However, POT approach remains underemployed in Australia.…”